Two posts, so far
Writing
Writing about measuring AI answer engines: what ChatGPT and Gemini actually say
when a buyer asks for the best option in a category, how often they contradict
themselves, and what that means for anyone selling AI visibility. Including the
findings we withdrew and the numbers we corrected in public.
Same mechanism, different mediator
SEO is a discovery mechanism applied to search engines. GEO is the same mechanism applied to language models — not a new system, and the difference decides what survives the next shift.
We measured Gemini as 3× less stable than ChatGPT. Then we tripled the sample.
Our first instability figure said Gemini was about 3.5× noisier than ChatGPT. At triple the sample it was 1.6×. What moved, what conclusion it killed, and why the correction is published rather than swapped.